/Scored2.0

A neural network to predict cricket scores and win percentage.

Primary LanguageJupyter Notebook

Scored2.0


Description

Scored 2.0 is a collaborative project undertaken by three passionate individuals, Muhammad Arsalan Khan, Athar Rizwan, and Aun Noman. Our mission? To predict cricket scores and win percentages using innovative machine learning models.

The Code Behind It

While the front end of our project may seem simple, the magic happens behind the scenes. If you're curious to explore the inner workings and dive into the code, follow these paths:

The project structure is organized into three main folders, each named after one of our team members: Aun, Athar, and Arsalan. Despite our collaborative efforts, we deliberately worked independently to bring diverse perspectives to the table. This is most evident in the distinct implementations of Arsalan's and Athar's models, particularly when handling wide and extra balls.

In essence, each team member's folder contains three key subfolders:

  • Notebooks
    • Here, you'll find all the code—both for data collection and augmentation, as well as model development.
  • Resources
    • This is where we store our data in pickle and CSV formats, after augmenting it in the notebooks.
    • IMPORTANT! The Resources folder may not be present in each team member's directory due to its size. You can access these resources via the provided links:
  • Models
    • This is the home of our trained final models, which are utilized to showcase the magic in action.

Contributing

We're passionate about advancing our project, and we firmly believe that there's more to explore and enhance. If you're interested in contributing, we welcome collaboration.

Here are some ideas we're excited to explore further (and we'll be working on them ourselves too):

  • Deploy the model for wider use.
  • Develop a user-friendly frontend website or app to support our model.
  • Incorporate additional data and features for even more accurate predictions.
  • Experiment with different types of models for further improvements.

Contact Us

Feel free to reach out to us via email:

Let's connect and make cricket prediction more exciting and accurate together! 🏏🤖📈